This invention relates generally to Coriolis effect transducers and more particularly to test, or calibration, structures for such transducers.
As is known in the art, Coriolis effect transducers are used to measure angular rate. More particularly, if a mass, m, with velocity, v, relative to the body of the transducer, experiences an angular rate xcexa9IN about a rate sensing axis perpendicular to the velocity of the mass, the mass will experience a Coriolis acceleration, ACOR=2vxcexa9IN, along a sensitive axis perpendicular to both the velocity, v, and the rate sensing axis. Thus, a Coriolis force FCOR is produced on the mass equal to mACOR along the sensitive axis. The motion of the mass is restrained in the direction of the Coriolis force by a mechanical or electrical restraint, such as a mechanical spring or electrical servomechanism. The mass will undergo a displacement y=FCOR/k, along the sensitive axis, where k is a constant, such as the spring constant or reciprocal of the spring compliance. This displacement y may be measured by any displacement transducer, such as, for example, a device which measures difference in capacitance produced by a linear change in a gap (i.e., displacement) between plates of a capacitor, one of such plates being mechanically coupled to the mass and the other being fixed relative to the body of the transducer. The device then produces an electrical output signal VOUT proportional to this change in capacitance. Thus, such a Coriolis transducer produces an output electrical signal, VOUT=Kxcexa9IN, where K is a proportionality constant which is a function of the physical and electrical properties of the transducer.
In order to determine K, a calibration procedure is used. Such calibration procedure typically includes setting the transducer on a rate table and applying a known input angular rate, xcexa9IN TEST, about the rate sensing axis while the output signal VOUT TEST is measured. The proportionality constant, K, is determined in accordance with K=VOUT TEST/xcexa9IN TEST. While this calibration procedure provides a determination of K, it is time consuming and therefore adds cost to the transducer.
As is also known in the art, one technique used to provide relatively inexpensive Coriolis transducers is micromachining. One such micromachined Coriolis transducer is described in my U.S. Pat. No. 5,635,640 issued Jun. 6, 1995, the entire subject matter thereof being incorporated into this patent application.
It is also very useful to be able to verify the quantitative functionality of a packaged Coriolis sensor in its end use. Specifically, when such sensors are used for automotive rollover or anti-skid applications they become safety-critical items and it enhances their usefulness if their correct function can be established with high confidence every time they are used. As will be described, this invention provides a means of testing that function with high confidence.
In accordance with the present invention, a method is provided for testing a Coriolis transducer having a body with a mass, m, vibrate along a vibratory direction in a resonant structure and undergo a displacement along a sensitive axis in response to an angular rate about a rate sensing axis. The displacement is perpendicular to both the vibration and the rate sensing axis. The method includes: applying forces, FTEST VIBRATORY and FTEST SENSITIVE, on the mass along the direction of vibration and along the sensitive axis, respectively, in a predetermined ratio, N; and, measuring an incremental output VOUT TEST of the transducer in response to the forces, FTEST VIBRATORY and FTEST SENSITIVE. The FTEST vibratory is identical to the vibratory drive force in normal operation.
With such method, a known test input angular rate xcexa9IN TEST SIM is simulated, such rate, xcexa9IN TEST SIM, being a known function of a measured characteristic frequency of the resonant structure. Thus, the proportionality constant K=VOUT/xcexa9IN=VOUT TEST/xcexa9IN TEST SIM can be calculated without expensive rate table testing. Further, because the test can be performed with a transducer prior to packaging, electronics integrally formed on the same semiconductor wafer of the mechanical transduction structure may be easily trimmed to provide a desired proportionality constant K.
In accordance with another feature of the invention, the mass of the transducer is included in a resonant structure. During a test, or calibration, mode, with the forces, FTEST VIBRATORY and FTEST SENSITIVE having a frequency at, or near the resonant frequency of the mass, and in the absence of an input angular rate about the rate sensing axis, the velocity of the mass will be predominately FTEST VIBRATORY/D, where D is the damping factor of the body within the resonant structure. Thus, under such condition, a simulated input angular rate, xcexa9IN TEST SIM={FTEST SENSITIVE/FTEST VIBRATORY}{BW/2}, where BW is the resonant bandwidth of the resonance structure is applied to the transducer. In response to such simulated input angular rate, xcexa9IN TEST SIM={FTEST SENSITIVE/FTEST VIBRATORY}{BW/2}, the output of the transducer VOUT TEST will be K{BW}/2N; i.e., K=2N{VOUT TEST)/BW. Because the mechanism used to apply the vibratory force, FTEST VIBRATORY, on the mass along the vibratory direction and the mechanism used to apply the force to the mass along the sensitive axis, FTEST SENSITIVE, are fabricated in proximate regions of the transducer, they have matched physical and electrical characteristics. Thus, manufacturing variations incurred in the formation of one of the force mechanisms occur to the other one of the force mechanisms with their size ratio, N, being independent of such manufacturing variations. As a consequence, with the vibratory direction force to sensitive axis force ratio, N=FTEST VIBRATORY/FTEST SENSITIVE, being related to the ratio of the sizes of the mechanisms, rather than to the absolute size of each one of the mechanisms, such ratio, N, can be accurately fixed by the design of the structure. In short, the ratio N is independent of manufacturing tolerance. The remaining parameter, the resonance bandwidth, BW, of the resonant structure, is readily and quickly ascertainable from a frequency response measurement of the resonant structure. While the resonance bandwidth, BW, changes with manufacturing variations, it is measurable from a frequency response measurement which does not require application of a known input angular rate to the transducer.
In one embodiment, the vibratory direction force mechanism, FTEST VIBRATORY, during both the normal mode and the test mode, and the sensitive axis force mechanism, FTEST SENSITIVE, used during the test mode, are electro-statically driven fingers. Thus, N is merely the ratio of the number of fingers used to produce FTEST SENSITIVE to the number of fingers used to produce FTEST VIBRATORY. Consequently, an accurate determination of the proportionality constant, K is achieved.
In accordance with another embodiment of the invention, during the test mode, the forces FTEST SENSITIVE and FTEST VIBRATORY are applied with a frequency, xcfx89APPLIED, less than the natural resonant frequency, xcfx89o. In such case, the simulated xcexa9IN TEST SIM={xcfx89o2}/{2Nxcfx89APPLIED}, in which case the natural frequency xcfx89o is measured and used for the calibration (i.e., the characteristic frequency of the resonant structure measured and used for the calibration is the natural frequency xcfx89o). In particular, the proportionality constant K=VTEST OUT/xcexa9IN TEST SIM={2Nxcfx89APPLIED/xcexa9o2}{VTEST OUT} is determined when the vibratory direction and sensitive axis forces, FTEST VIBRATORY and FTEST SENSITIVE, respectively, are applied to the transducer in ratio, N, with xcexa9IN=0, and VTEST OUT is measured at the output of the transducer.
In another embodiment of the invention, during the test mode, the forces FTEST SENSITIVE and FTEST VIBRATORY are applied with a frequency, xcfx89APPLIED, greater than the natural resonant frequency, xcfx89o, the simulated xcexa9IN TEST SIM=xcfx89APPLIED/2N, in which case the characteristic frequency used for the calibration is xcfx89APPLIED. In particular, the proportionality constant K={2N/xcfx89APPLIED}{VTEST OUT} is determined when the sensitive direction and the vibratory axis forces, FTEST VIBRATORY and FTEST SENSITIVE, respectively, are applied in ratio, R, to the transducer, with xcexa9IN=0, and VTEST OUT is measured at the output of the transducer.
In accordance with another aspect of the invention the vibratory direction force mechanism and the sensitive axis force mechanism are in proximate regions of the transducer.
In accordance with still another aspect of the invention, the vibratory direction force mechanism the sensitive axis force mechanism have matched physical and electrical characteristics.
In accordance with another aspect of the invention, a Coriolis transducer is provided. The transducer includes a resonant structure having a mass vibrate along a vibratory direction in a resonant structure and undergo a displacement along a sensitive axis, perpendicular to the vibration, in response to an angular rate about a rate sensing axis, such displacement being perpendicular to both the vibration and the rate sensing axis. The transducer includes a vibratory direction force mechanism adapted to apply a vibratory direction force, FTEST VIBRATORY, on the mass along the vibratory direction during both a normal operating mode and during a test mode. Also provided is a sensitive axis force mechanism adapted to apply simultaneously to the vibratory direction force, a force, FTEST SENSITIVE, to the mass along the sensitive axis when such sensitive axis force mechanism is coupled to the mass. A switch is included to couple the sensitive axis force mechanism to the mass during a test mode and being adapted to de-couple the sensitive axis force mechanism from the mass during a normal operating mode.